Grano Colloquium

January 16
Harper 130
University of Maryland
Control without finiteness contrasts: Case and complement size in Mandarin Chinese

A fundamental question for control theory concerns the distribution of controlled arguments. The restriction of control to nonfinite environments in English was built deeply into early and GB-era theories of control (Chomsky 1981), but in recent years, theories have had to countenance an increasingly rich crosslinguistic range of data both in the nuances of the finite/nonfinite distinction and its interaction with the distribution of control: among the most well studied cases are control into some finite subjunctive complements in Balkan languages (Terzi 1992; Iatridou 1993; Landau 2004) and control into finite indicative complements in Brazilian Portuguese (Ferreira 2000; Rodrigues 2002). In contrast, relatively little attention has been paid to the distribution of control in languages that lack overt tense and agreement morphology. One such language, Mandarin Chinese, has been argued by some to have a covert finite/nonfinite distinction responsible for the distribution of control (Huang 1982; Li 1990) and by others to lack such a distinction and concomitantly not have syntactic control (Hu et al. 2001). In this talk, I argue that Mandarin does have syntactic control but does not have a finite/nonfinite distinction. Instead, based in large part on a detailed investigation of the behavior of aspect marking in complement clauses, I propose that controlled complements in Mandarin are bare vPs and consequently require control because they lack the functional structure needed to Case-license an overt DP. The upshot is that some version of the Case-based approach to the distribution of control is correct (Bouchard 1984) and that some of the burden of explaining the distribution of control can be shifted away from distinctions like (non)finiteness — which are notoriously difficult to justify in Chinese-like languages  — and onto splits in complement size — splits made available in some sense ‘for free’ by the architecture of the clause and also well documented for other languages under the name ‘restructuring’ or ‘clause union’ (Wurmbrand 2001; Cinque 2006).